Recent attempts at treating Alzheimer's patients with a beta-amyloid vaccine have resulted in an adverse brain inflammation response for a significant number in the trial group. The full-length 42 amino-acid beta-amyloid peptide was used as a vaccine in that study. Given the size of that molecule, it is expected to present many potential epitopes against which antibodies can be generated. It is hypothesized that antibody interaction with one or a few of those epitopes would be therapeutically beneficial while binding to other sites could have resulted in the unfortunate side effects. The proposed research will define which sites on the molecule are best to target therapeutically and which are best to avoid so that any untoward side effects are circumvented. The immunotherapy of Alzheimer's disease will become sharply focused and more consistent by using epitope specific anti-beta-amyloid scFv human antibodies for passive immunization. A large panel of beta-amyloid antigens that represent different sites on the beta-amyloid molecule has been synthesized. These will be used to screen a yeast recombinatorial human single chain Fv antibody display library of over one billion different antibodies. Unique clones, which tightly bind to specific epitopes on the beta-amyloid molecule, will be identified. Yeast expressing those select anti-beta-amyloid antibodies will be induced and the secreted single chain Fv purified via its his tag. Several precise binding assays will be used to characterize the specificity and affinity of each distinct antibody. Functional assays will be used to test the ability of the different single chain Fv antibodies to dissolve beta-amyloid aggregates in vitro. Therapeutic potential will be evaluated using a transgenic mouse model. Effective treatment will be assessed by an ability to prevent cerebral plaque formation and/or dissipate preestablished plaques in the brain. Lastly this yeast display system will provide a unique opportunity to refine those therapeutic agents through genetic engineering so they become precisely tuned to the task at hand. For example the single chain antibodies can be modified to achieve higher binding affinities, greater functionality, bivalent binding, a prolonged serum half-life or passage into the CNS.